Ilmenau University of Technology Media Bias in
Transcrição
Ilmenau University of Technology Media Bias in
Ilmenau University of Technology Institute of Economics ________________________________________________________ Ilmenau Economics Discussion Papers, Vol. 20, No. 99 Media Bias in Women´s Magazines: Do Advertisements Influence Editorial Content? Julia Pannicke December 2015 Institute of Economics Ehrenbergstraße 29 Ernst-Abbe-Zentrum D-98 684 Ilmenau Phone Fax 03677/69-4030/-4032 03677/69-4203 http://www.wirtschaft.tu-ilmenau.de ISSN 0949-3859 Media Bias in Women´s Magazines: Do Advertisements Influence Editorial Contents? Julia Pannicke# + Abstract: Existing theoretical and empirical studies on Media Bias are subjects of controversial discussions in the literature. However, scientific literatures on Media Bias establish empirical evidence for a positive impact of advertising volume on media coverage. To get in line with the debated literature about whether biases occur, this economic paper presents an empirical analysis of a possible (commercial) Media Bias influenced by advertising expenditure in monthly women’s magazines. The results of a linear panel model regression, a panel poisson regression, as well as those of a panel negative binomial regression model show that there is a positive correlation with the amount of advertising expenditure on the coverage of a company that purchased advertisements on the women’s magazines in Germany. A positive correlation between advertising volume and the nomination as well as prize winning of (cosmetic) products could also be found. Keywords: Media Bias, Advertisement, advertising volume, coverage, two-sided markets # + M.Sc., Institute of Economics, Ilmenau University of Technology, Email: [email protected]. This paper is based upon a data set that was predominantly collected by Julia Walter in the con text of her master thesis (Walter 2014). I thank Julia Walter a lot for providing and explaining her data. 1 1. Introduction Media markets represent a prime example of a two-sided market (see inter alia Anderson and Gabszewicz (2006), Dewenter (2006)), whereas recipients are offered entertaining and informative content, while advertising customers spent their budget in order to reach a potential high number of recipients. For media companies, especially for print media, advertising is the key and unique source of profits (Dewenter and Heimeshoff (2014), Bralso and Sobbrio (2011)) to remain lucrative. Against this background the question arises whether magazines or newspapers have major incentives in distorting editorial contents for the benefit of its own advertising customers to increase their demand for placing advertisements. While free press in a democratic constitutional state shall preserved unbiased coverage and plurality of opinions, various studies have for example shown that news reports are on any number of occasions biased in a special political direction (inter alia Mullainathan and Shleifer (2005); Gentzkow and Shapiro (2006); Chan and Suen (2008); Sobbrio (2011)). This is where the young research field of Media Bias comes into its own. This field is located at the interface of several scientific disciplines and took economists, political scientists as well as media and communication scientists to academic discussions. Nevertheless, most of the various researchers have recently focused their study on Media Bias in the field of political, media and communication science: the relationship between advertisers and recipients, however, is hitherto far less discussed in the literature. All in all, there are three different theoretical approaches aimed at explaining the existence of biases (Dewenter 2011:12). I) A random and unsystematic distortion of coverage is the least problematic type of media bias. In such a case, the distortion can be simply explained by the fact of less precise and incorrect journalistic research work. II) The ideological distortion, however, is done on purpose in order to give a false impression of reality. This kind of distortion may be politically, racially, religiously or socially motivated and is often not-profit oriented (see inter alia Mullainathan and Shleifer (2005); Gentzkow and Shapiro (2006 and 2010)). III) The last approach deals with the type of profit oriented Bias, particularly through adaption to the recipient’s preferences (see inter alia Hamilton (2004); Ellman and Germano 2 (2009)). This type of media bias with particular having regard to platforms of a twosided market (Rochet and Tirole (2003)) is of importance for this paper. The incentives for distorting coverage are on the one hand with respect to the recipient’s preferences or on the other hand to the advertiser´s preferences, which may cause an increasing demand on the readers or advertising market. Thus, there is a direct or indirect way of maximizing the revenue on the market for advertising. This is especially problematic when it comes due to customizing to the advertiser´s preferences to a strong biased news coverage and consequently to a unilateral coverage and reporting style. In conclusion, distortions lead to an insufficient provision of information, an increase of information costs and therefore to welfare losses (Dewenter (2011)). Furthermore, distorted content in newspapers or magazines appear to be “objective” and therefore the effect of persuasive advertising and the power to convince consumers increases (Gambaro and Puglisi (2010):1). Theoretical studies on media bias are ambiguous, arguments for and against media biases are discussed in the literature. In conclusion, it is an empirical question that is already focused in research. At this point, I follow the existing discussion and complete the economic literature by an empirical study on Media Bias. Because only a small number of studies have recently concentrated on the effects of advertising expenditure on coverage, this paper deals with the topic of commercial Media Bias. For that reason, I focus on a possible link between advertising expenditure and media coverage in the field of German woman´s magazines. Indeed, the market for women's magazines is characterized as a two-sided market, which means, that the market for advertising and the market for recipients are dependent from each other (see Bagwell, 2007). This raises the question whether magazines seem to bias their coverage towards their main advertising customers in order to “reward” their advertising costumers. More concretely, I investigate the relationship between the yearly number of mentions of companies in the editorial sections and the yearly amount of advertising that were purchased by these companies on two woman´s magazines during the period 2009-2012. Consequently, the interesting research question is: whether the advertising volumes of a company in the selected magazines increase 3 the absolute number of mentions in the coverage of its products. In conclusion, I find a positive significantly correlation between the advertising expenditure and the mentions in the magazine’s coverage of these companies that purchased their advertisements. The paper is arranged as follows. Section 2 briefly focuses some earlier studies concentrating on Media Biases and especially the relationship between advertising expenditure and coverage of newspapers or magazines. Section 3 forms the main part, determining backgrounds of the investigated magazines, containing a data description of the panel data model, the econometric estimation and controlling variables. Section 4 and 5 discuss the results of the econometric analysis as well as summarizes and concludes. 2. Literature Review Over the past few years, both theoretical and empirical studies have been focusing on the influence of advertising on reporting. Gal-Or et al. (2010) extend the theoretical study of Mullainathan and Shleifer (2005), dealing with political bias, by analyzing a biased coverage when advertising customers play a substantial role in terms of revenue. Gal-Or et al. (2010) show that (commercial) Media Bias depend on the concept of single or multihoming of advertising customers. Multihoming and singlehoming means whether a company advertises in more than one newspaper (or only one newspaper or magazine) and therefore chooses to share (or nor to share) the advertising budget. Accordingly, if the concept of singlehoming is followed, a tendency of distortion in favor of advertisers can be shown. Gal-Or et al. (2010) cannot find a distortion tendency in case of multihoming. Ellman and Germano (2009) investigate to what extent the market structure effects a commercial media bias in a two-sided market for newspapers. Here, they find that advertising expenditure influences news coverage only in monopolistic newspaper markets. On the contrary, competitive newspaper markets lead to, although the amount of advertisements increases, unbiased news coverage. Blasco and Sobbrio (2011) likewise investigate the relationship between competition and commercial 4 media bias also with regard to the market structure of the advertising customers. They introduce a model by combining the central and important statements of the theoretical literature. In the end, Blasco and Sobbrio (2011) formulate three central aspects for media regulators that should be considered regarding commercial media bias. The literatures that are closest to this research are the empirical ones on advertising and media coverage by Reuter and Zitzewitz (2006), Reuter (2009), Dewenter and Heimeshoff (2014), Gentzkow, Glaeser and Goldin (2006)), Poitras and Sutter (2009), Rinallo and Basuroy (2009) ,and Gambaro and Puglisi (2010). Reuter and Zitzewitz (2006) investigate the relationship between advertising expenditure and mutual fund recommendations in financial magazines and financial parts of national newspapers. Through their fixed effects logit model and controlling for characteristics of funds and other confounding variables, Reuter and Zitzewitz (2006) find a positive relation between mutual fund recommendation and past advertising expenditures. The latter, however, does not apply to all finance magazines but affects three magazines (Money Magazine, Kiplinger's Personal Finance and Smart Money), who specialize in reporting on personal investing. The national newspapers (Wall Street Journal and New York Times) that are lesser financed by advertisement of mutual funds, show no positive correlation between advertising expenditure and positive mentions in the magazines. Another well-known study by Reuter (2009) pursues the issue whether there is a correlation between wines’ ratings and advertising expenditure in a wine magazine. He finds weak evidence for biased ratings. Dewenter and Heimeshoff (2014) deal with the question of up to what extent advertising expenditure of automobile manufactures’ impact the test score of the German bi-weekly car magazines Auto, Motor und Sport. Compared to the study of Reuter (2009), which investigates the effects of advertising volumes on occasionally subjective wine ratings in some sense, Dewenter and Heimeshoff (2014) benefit from measuring the products performance in an objective way. Thus, in order to explain the 5 test results, the product performance of a car can be controlled through technical characteristics such as horsepower, mileage, diesel engine, the number of doors or if the car is constructed by a German company. Beyond that, they also collect advertising volumes of the respective car manufacturers to analyze a possible relationship between the advertising expenditure and the performance of the respective cars in the ranking. By using a two-step procedure and accounting for endogeneity and sample selection (probability of a car to be refereed at all) they find a distortion on test scores in favor of advertisers. In most of their cases, they find a selection bias which means that it is more probable to be reviewed when the advertising volume is high. In addition to this, quite a few authors likewise research commercial media bias by concentrating on the US newspaper industry (Gentzkow, Glaeser, and Goldin (2006)), muckraking newspaper (Poitras and Sutter (2009)), fashion firms in magazines and newspapers (Rinallo and Basuroy (2009)) as well as Italian listed companies in daily newspapers (Gambaro and Puglisi (2010)) or the influence of bargaining power of advertisers (Guo and Lai (2014)). 3. Econometric Analyses 3.1 Data description This paper investigates the relationship between advertising volume and distorting coverage. Subject matters of the analysis are monthly women's magazines. This is due to several reasons. First of all, the gross advertising expenditure in women's magazines is generally very high. According to Nielsen Media Research, monthly woman’s magazines spend an amount of 348,5 million euros in 2012 and an even larger amount of 378,1 million euros in 2013 for advertising. Thus, the gross advertising expenditure of monthly woman’s magazines hold the second place out of all consumer magazines in Germany (Nielsen). Furthermore, woman’s magazines are among the top five of the most popular types of magazines in Germany. 35 % of Germans that are older than 14 years read magazines for women (VuMA). 6 While advertisements in newspapers are more likely to be regarded as an annoying effect, advertisements in women's magazines are rather perceived as friendly and positive (News Aktuell (2008), Brigitte (2010, 2012)). Kaiser and Song (2009) collected data of different consumer magazines in Germany and show through their logit demand estimation that reader´s even esteem advertisements. Interestingly, advertisements are besides advertisements in TV magazines the most appreciated in women´s magazines. Additionally, printed advertisings in magazines are usually a very accepted form of advertising because of the special trusting relationship. This can be explained by the fact that consumer magazines can convey the impression of being an expert-magazine in their field, especially in the area of "special interest"- magazines. That is exactly the case of woman’s magazines. In this way, the selected advertisings enjoy a high credibility right from the start (VDZ, McKinsey (2012)). This special trusting relationship may also be assumed and transferred to the coverage of the printed papers. Furthermore, the frequently discussed crisis about printed media in the field of consumer publications has no apparent impact on the number of existing magazine titles (Jarren (2010), Diez (2009)). On the contrary, the number of published consumer magazines has been increasing from 1,048 to 1,587 titles in Germany within the period from 1997 to 2014 (VDZ (2014), Statista (2015)). For this paper, the highest-circulation and the highest-reach of women's magazine were chosen. Using data from the `Informationsgemeinschaft zur Feststellung der Verbreitung von Werbeträgern e. V.‘ (short: IVW), the Glamour magazine is the highest-circulation magazine within the time span 2009-2012 (IVW (2014a)). For the purpose of analyzing the highest-reach magazine the data of „Media-Analyse agma Media-Micro-Census“ were evaluated. The highest-reach women’s magazine within the investigated time period is the Cosmopolitan magazine (Walther, 2014). In summary, Glamour and Cosmopolitan were chosen within the period from 01/2009 to 12/2012 as monthly women's magazines at national level. 7 Woman´s magazines The Cosmopolitan is an international monthly magazine for women, which was launched by Schlicht & Field of New York in 1886. Published in Germany since 1980, the Cosmopolitan has a worldwide distribution in around 100 countries and has 64 editions and is published in 35 different languages. In 1886, the magazine was taken over by the Bauer Media Group, previously it belonged to Marquard Media AG. In Germany, the Cosmopolitan is published in a pocket (A5) and a classic (A4) format, the latter was used for this study. Besides the investigated printed versions, the cosmopolitan is also available in digital formats (e.g. website, variety of social media). Of particular noteworthiness is also the annual presentation of the "Prix de Beautés" as a price of special products regarding the market of cosmetic products. The price for particularly innovative products is chosen from around 600 products and is annually awarded by judges of experts consisting of dermatologists, editors and industry representatives. The awarded products in 15 different cosmetic categories are presented within the printed version usually in March. This presentation should be classified not as advertising but rather as editorial content of the magazine. The Cosmopolitan achieved average ratings of 1,8 million readers per issue (Allensbacher market Media Analysis, AWA (2014)) and appears 12 times annually in an average print run of 444,023 copies per quarter during the period 2009 to 2012 (IVW (2014a)). The typical cosmopolitan reader is between 20 and 49 years old, employed, and has an above-average household net income (Bauer Media Group (2014)). Similarly, Glamour is a monthly women's magazine, published by Condé Nast Publications. The first edition, called Glamour of Hollywood, was published in the United States in 1939. Since 2001, the magazine for women is also published in the Germanspeaking area (see Wilhelm-Fischer (2004) and Condé Nast Verlag (2014a). The magazine is distributed in 15 international core markets (Condé Nast Verlag (2014)). In Germany, the magazine is exclusively printed in pocket format (DIN A5 format). In addition to the printed editions, the magazine can be found in various social networks and its own website (Condé Nast Verlag (2014)). As with the cosmopolitan, only the printed edition is substantial for the investigation. 8 In this context, of particular interest is an annual awarded cosmetic-prize called “Glammy” (Condé Nast Verlag (2012)) that is selected by the readers from a pool of nominees. In other words, Glamour´s readers are asked to choose their favorite products out of three nominated products from various categories. The three available products were nominated by the editors beforehand. Besides the possibility of voting via internet, the nominated products are presented in each February issue and the winners are presented one month later (Condé Nast Verlag (2012)). A further characteristic of the Glamour is the so called shopping-week, which is organized twice a year with a special offer of discounts. The shopping-week pages are clearly separated from the rest of the magazine’s content. Glamour achieved average ratings of 1,56 million readers per issue in 2014 (AWA (2014)) and also appears 12 times a year. The average print run was 660,361 copies per quarter during the period 2009 to 2012 (IVW (2014b)). The target group of the magazine are 33 years old women on average, educated, employed, and with an average household net income of € 3,127 monthly. They are interested in shopping, exploring new trends, and are openminded (Condé Nast Verlag (2014), S.11). All in all, the typical readers of these magazines are very similar to each other and both of the magazines award yearly prizes for the best products. 3.2 Data collection and descriptive statistics To estimate the influence of companies’ advertising volumes on the coverage of its products in woman’s magazines, data on mentioning of companies in editorial contents of the magazines were collected as well as a proxy for advertising volume. Furthermore, control variables have to be considered. Advertising expenditure At first, all advertisings in all issues from 01/2009 to 12/2012 were counted and summarized for both magazines. Altogether, the following parameters for each advertisement were set: issue, brand, product, sector, sub-sector, advertising size, advertising position, product samples, front/back cover and special format/special paper. 9 In general, the Cosmopolitan has 197 pages on average, 47 % of the pages are filled with advertisements. The Glamour counts 239 pages on average with advertising content of 48 %. All in all, a total of 3.449 advertisements were captured in the Cosmopolitan and a total of 4,131 in the Glamour magazine. The Cosmopolitan included 1,565 ads in the sector of cosmetic products, followed by the fashion industry with 545 and the media sector with 385 ads during the investigated period. As you can see in figure 1, 45 % of advertisements are due to the cosmetic sector. Figure 1: Percentage of Advertisments per sector, Cosmopolitan Food 3% Technics 4% Others 15% Cosmetic 45% Jewerly 6% Media 11% Fashion 16% Others Food Technics Jewerly Media Fashion Cosmetic The Glamor counted 1,359 ads in the sector of cosmetic, 1,265 of the fashion industry and 281 of the media sector. Figure 2 shows the relative distribution by the different sectors. The largest advertising-sector is the cosmetic industry for both magazines. Therefore, cosmetic products and companies that purchased their advertisements in the field of cosmetic are considered in this study. 10 Figure 2: Percentage of Advertisments per sector, Glamour Others 14% Gastronomy 2% Jewerly 6% Cosmetic 33% Onlineshop 7% Fashion 31% Media 7% Others Gastronomy Jewerly Onlineshop Media Fashion Cosmetic . To approximately determine the advertising expenditure for the cosmetic sector, 2,924 advertisings were collected for the period of observations in total. Furthermore, 749 different companies purchased these advertisements in both magazines. In order to measure the advertising expenditure, a proxy variable has to be generated. The advertising volume of each company can simply be calculated. In order to do so, some variables have to be taken under consideration 1. The first dummy variable for calculating the advertising volume is a variable that indicates whether the ad was placed before page 41 of the Cosmopolitan magazine or not. This information is important because advertising is more expensive before reaching page 41 (Bauer Media Group (2013)). For Glamour it does not make any difference in price calculation to advertise before a special number of pages. Because of the different prices, another two dummy variables has to be taken in consideration. These are a dummy for opening spread and a dummy for the back cover. The variable opening spread is 1, if advertisement is presented on page 2 to 3, otherwise equals 0. The dummy variable back cover equals 1, if advertisement is presented outside back cover, otherwise equals 0. 1 Of course, not every data could be gathered, such as parameters for special discount offers and product samples as well as added advertising material. 11 The following variables are also relevant because of their differences in prices. The dummy variables SW describe whether the ad appeared in the part of Glamour "shopping week". Another Dummy CGF is an indication of whether the advertiser has booked two more pages beyond the opening spread pages. The variable IBC stands for the inside back cover. These variables play only a relevant role for calculating the advertising volume for the Glamour magazine, also the shopping week variable. The size of the advertisement plays also a relevant role for calculation (e.g. full or half page, besides the content, double page). The following equation is used for calculation the advertising volume per issue and per company’s advertising for cosmopolitan magazine: 𝐴𝐴𝐴𝐴 𝑖𝑖𝑖𝑖𝑖𝑖 = (𝑃𝑃𝑖𝑖𝑖𝑖≤41 𝑥𝑥 𝑅𝑅𝑖𝑖𝑖𝑖 ≤41 + 𝑃𝑃𝑖𝑖𝑖𝑖>41 𝑥𝑥 𝑅𝑅𝑖𝑖𝑖𝑖 >41 + 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑥𝑥 𝑅𝑅𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 + 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑥𝑥 𝑅𝑅𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 ) − 𝐷𝐷𝑖𝑖𝑖𝑖 𝐴𝐴𝐴𝐴 𝑖𝑖𝑖𝑖𝑖𝑖 = 𝑃𝑃𝑖𝑖𝑖𝑖≤41 = 𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑃𝑃𝑖𝑖𝑖𝑖>41 = 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 𝐷𝐷𝑖𝑖𝑖𝑖 = (1) Advertising volume AV of company C for the issue i of Cosmpolitan C Number of pages P including page 41 Advertising rate R Number of pages P starting on page 42 Opening spread OS, advertisement is presented on page 2 to 3 Advertisement is presented outside back cover BC Discount D, if a company booked a special amount of pages Similarly, the advertising volume calculation regarding the Glamour magazine, with exception of some variables (e.g. page 1 to 41): 𝐴𝐴𝐴𝐴 𝑖𝑖𝑖𝑖𝑖𝑖 = (𝑃𝑃𝑖𝑖𝑖𝑖 𝑥𝑥 𝑅𝑅𝑖𝑖𝑖𝑖 + 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑥𝑥 𝑅𝑅𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 + 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑥𝑥 𝑅𝑅𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 + 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑥𝑥 𝑅𝑅𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 + 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑥𝑥 𝑅𝑅𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 + 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 𝑥𝑥 𝑅𝑅𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 ) − 𝐷𝐷𝑖𝑖𝑖𝑖 𝐴𝐴𝐴𝐴 𝑖𝑖𝑖𝑖𝑖𝑖 = 𝑃𝑃𝑖𝑖𝑖𝑖 = 𝑅𝑅𝑖𝑖𝑖𝑖 = 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 𝑃𝑃𝑖𝑖𝑖𝑖𝑖𝑖𝑖𝑖 = 𝐷𝐷𝑖𝑖𝑖𝑖 = 12 Advertising volume AV of company C for the issue i of Glamour G Number of pages P Advertising rate R Advertisement is presented on cover gate fold CGF Opening spread OS Advertisement is presented inside back cover IBC Advertisement is presented outside back cover BC Advertisement is presented in a shopping week SW issue Discount D, if a company booked a special amount of pages (2) For calculating the advertising volume, the price lists for both magazines were used of the year 2013. While the advertising prices during the time period from 2009 to 2013 have continues increased in total, the differences between the advertising volumes of several companies were almost unchanged. Therefore the price list of only on single year can be used without hesitation (see Bauer Media Group (2013)). Mentions For the purpose of determining the relationship of mentions and advertising volume, all mentions of the cosmetic sector were collected that appeared in the editorial content of both magazines. Consequently, all mentions of companies’ products in the editorial part were counted by hand. Again parameters have been identified regarding the mentions of a product of a certain company, such as issue, page, year, magazine and the products’ mentioning in a continuous text or as an image of the coverage. Altogether, 3,153 mentions of companies’ products were counted in the Cosmopolitan and 3.258 in the Glamour magazine within the investigated period of time. Descriptive Statistics In order to get a first impression on potential biased coverage and the structure of data, I look at the top 10 companies with the highest advertising volumes as well as the products with the most mentions over the entire investigated time period of several companies. The descriptive statistics can be found in the following tables. While table 1 shows the top 10 companies of Cosmopolitan magazine, table 3 illustrate those 10, who were most frequently mentioned and those with the highest advertising volume in the Glamour magazine. 13 Table 1: Cosmopolitan 2009-2012, Advertising Volume (AdV) and Mentions in total per company Ranking Company AdV in Euro Company 1 2 3 4 5 6 7 8 9 10 L’Oréal Nivea Maybelline Lancôme John Frieda Dior Vichy Artdeco Clinique Estée Lauder 6323460 3877407 2925220 2892055 2725225 2647170 2440475 2160750 2043500 1809000 L’Oréal Dior Chanel Lancôme Artdeco Nivea Clinique Estée Lauder Clarins Shiseido Mentions 132 132 109 98 96 94 87 87 87 82 It can be seen that 7 out of 10 companies, that were most frequently mentioned, are in the top 10 of the largest advertising customers in the Cosmopolitan. Table 2: Cosmopolitan 2009-2012, Affiliated Group: Advertising Volume (AdV) and Mentions in total Ranking Concern AdV in Euro Concern Mentions 1 2 3 4 5 6 7 8 9 10 L’Oréal Coty Beiersdorf Kao / Kanebo Procter & Gamble Estée Lauder LVMH Shiseido Co. Artdeco Cosmetic Gr. Unilever 20934820 5853455 5743022 5130525 4877265 4033400 3672270 3124210 2194250 1792585 L’Oréal Estée Lauder Coty LVMH Artdeco Cosmetic Gr. Procter & Gamble Beiersdorf Clarins Shiseido Co. Chanel 695 301 231 197 174 168 138 117 111 109 The numbers in table 2 show, that the affiliated group Coty (inter alia adidas, Celine Dion, Chloé, Kylie Minogue, Davidoff, Esprit, Jil Sander, Jette Joop) is achieving the ranking number two in spending money for advertising. This is a rather interesting outcome, since none of the brands mentioned in table one are part of the Coty Group. Now, the affiliated group Coty is also one of the top 3 concerns mentioned in the Cosmopolitan. As a result, further investigations on affiliated group based rankings should be part of the estimation. It can be seen that 8 out of 10 concerns, that were most frequently mentioned, are in the top 10 of the largest advertising customers in the Cosmopolitan. 14 Table 3: Glamour 2009-2012, Advertising Volume (AdV) and Mentions in total Ranking Company AdV in Euro Company 1 2 John Frieda 2704398 Dior 141 Nivea Dior Artdeco L’Oréal Clinique Douglas Garnier Guhl Yves Saint Laurent 2632982 2358024 2141300 2095128 1888855 1839645 1509926 1277300 1244160 Chanel Lancôme L’Oréal Yves Saint Laurent Clinique Clarins Nivea MAC Artdeco 108 101 98 91 90 87 70 68 66 3 4 5 6 7 8 9 10 Mentions Similarly, in the Glamour magazines are half of the mentioned companies in the top 10 of the largest advertising customers. Table 4: Glamour 2009-2012, Affiliated Group: Advertising Volume (AdV) and Mentions in total Ranking Concern AdV in Euro Concern Mentions 1 2 3 4 5 6 7 8 9 10 L’Oréal Procter & Gamble Coty Kao / Kanebo Beiersdorf Estée Lauder LVMH Artdeco Cosmetic Gr. Douglas Clarins 10532449 5385723 4014975 3981698 3375342 3226632 3045076 2365850 1871845 1629225 L’Oréal Estée Lauder LVMH Coty Procter & Gamble Clarins Beiersdorf Chanel Artdeco Cosmetic Gr. Kao / Kanebo 694 337 301 250 178 126 109 108 102 93 Underlying, also in respect of looking at the affiliated Group (table 4) in the Glamour magazine, only the affiliated group of Chanel is not one of the top 10 of the largest advertising customers in the Glamour. Chanel achieved the ranking number eight of the most frequently mentioned concerns. All other concerns are also one of the top 10 of the advertisers with the highest advertising volume in total. The top 10 rankings including both magazines are listed in Tables 5 and 6. Here you can find the total number of mentions and the top 10 of the largest advertising 15 customers (per company and per concern). Of course similarly, mentioned products of companies are often those, who offer a high amount of advertising. Table 5: Advertising Volume (AdV) and Mentions in total for both magazines Ranking Company AdV in Euro 1 2 3 4 5 6 7 8 9 10 L’Oréal Nivea John Frieda Dior Artdeco Maybelline Lancôme Clinique Vichy Douglas 8418588 6510389 5429623 5005194 4302050 4144617 3937255 3932355 3431627 3262045 Company Dior L’Oréal Chanel Lancôme Clinique Clarins Nivea Artdeco Yves Saint Lau. Estée Lauder Mentions 273 230 217 197 177 174 164 162 149 147 Table 6: Affiliated Group: Advertising Volume (AdV) and Mentions in total for both magazines Ranking Concern AdV in Euro Concern Mentions 1 2 3 4 5 6 7 8 9 10 L’Oréal Procter & Gamble Coty Beiersdorf Kao / Kanebo Estée Lauder LVMH Artdeco Cosmetic Gr. Shiseido Co. Douglas 31467269 10262988 9868430 9118364 9112223 7260032 6717346 4560100 3431870 3359245 L’Oréal Estée Lauder LVMH Coty Procter & Gamble Artdeco Cosmetic Gr. Beiersdorf Clarins Chanel Kao / Kanebo 1389 638 498 481 346 276 247 243 217 184 3. 3 Estimation Method All in all, 2,924 advertisings were collected for the period of observations in total. Moreover, 749 different companies purchased these advertisements in both magazines. Companies’ products were mentioned 6,411 times within the investigate time period. The goal is to explain a biased coverage through the unbalanced panel data set. Therefore, I use three different models calculating per year and per company- and 16 concern-level to check the robustness of the estimations. 2 (1) I run a linear panel model regression over time and over company, as well as over concerns. OLS regression panel model is used here because a continuous dependent variable is supposed. In order to model count variables (mentions), I used a (2) panel poisson regression over time and over company, as well as over concerns and a (3) panel negative binomial regression model that is usually used for over-dispersed count outcome variables with the same specifics (Cameron and Trivedi (1998)). All of these three models declaring advertising volumes as an explanatory variable. I focus on the number of mentions per year within the whole period from 20092012. I also focus those companies, that were not mentioned but purchased their advertisements and vice versa. Accordingly, I define 𝑚𝑚𝑡𝑡𝑡𝑡𝑡𝑡 as the dependent variable (3): 𝑚𝑚𝑐𝑐,𝑡𝑡 = α + βAdVc,t + γ + εc,t Where 𝑚𝑚𝑡𝑡𝑡𝑡𝑡𝑡 = AdVc,t = (3) the number of mentions of company c appearing on the women’s magazines (Glamour and Cosmopolitan) in year t, the corresponding advertising volume (monetary) by company c in year t, εc,t = error term and Xc,t includes corresponding control variables, such as a dummy variable if the product of a company is a brand that’s value is of one of the leading ten cosmetic brands worldwide3, the brand awareness by consumers of several cosmetic companies in % and a dummy variable among the fifteen most popular brands in the cosmetic sector. 4 These explanatory variables are used as a proxy for the “quality” or popularity of a company 5. 2 3 4 5 I do not draw a distinction between the magazines, because they resemble each other. Source: Millward Brown 2015, statista 2015. Source: Brigitte KommunikationsAnalyse 2012, S. 56, statista 2015. Because not every brand is considered, the sample size reduces when including these explanatory variables into the regression. 17 One major factor which concerns objective characteristics of the products needs to be taken into consideration. For instance, Dewenter and Heimeshoff (2014) achieve an objective measurement by including technical characteristics of car performance. Similar approach, unfortunately, cannot be easily used in this research paper by focusing on certain objective characteristics of cosmetic products and controlling for them. Quality is especially relevant for estimating the impact of advertising volume on the yearly awarded beauty products (“Glammy” and "Prix de Beautés") as well as the mentioning of a companies’ product. One source for this objective measurement of quality could be consumer ratings on internet platforms such as Ciao and Dooyoo or assessments made by independent institutions like Stiftung Warentest. In this case it is an obstacle, that there is either a biased rating caused by the amount of the product evaluations especially if certain products have not been considered at all. Because of this, it is not possible to control for quality in a direct way. Therefore, control variables such as the popularity or the value of a companies’ brand are considered. All in all, I estimate the equations (3) by all of the three models per company and per concern. Furthermore, after having tested for heteroscedasticity (Breusch–Pagan test), I estimate the linear panel model regression by means of robust standard errors. Awarded prizes I also estimated the relationship between the advertising expenditure and the awarded prizes and nominated products as a binary outcome variable. Because both magazines award a prize or respectivly a nominee for the best cosmetic products, the dependent variable can be defined as a nomination-dummy (for Glammy) as well as a winning-dummy of several companies within the investigated time-period. Because Glamour´s readers are asked to choose their favorite products out of three nominated products from various categories, the winners of the award do not play a role in the estimation. I only consider the nominated products by the editors beforehand. A probit regression model is used at this point. 18 4. Results and Discussion In general, the results from the linear model confirm the results of some empirical studies in the literature to the extent that the advertising volume has a positive effect on coverage. Table 7: Linear panel model regression per year and over company by means of robust estimators VARIABLES AdVc,t (1) Model 0.0000283*** (15.61) Valuable Brand (2) (3) (4) Model Model Model 0.0000272* 0.0000226* 0.0000293* ** ** ** (12.18) (11.54) (7.70) 3.409 (1.36) 4.985* (2.23) -0.247 (-0.05) 10.97*** (5.82) -9.616*** (-3.59) Brand popularity 0.230*** (4.03) Brand awareness % _cons N 2.093*** (16.60) 1416 2.097*** (16.62) 1416 1.995*** (17.11) 1416 4.192** (2.78) 145 t statistics in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 The estimation in column [1] of table 7 simply used advertising volume as an independent variable. The 10 most valuable brands for cosmetic products worldwide as a dummy variable were added in column [2] and a dummy if the advertised brand is a popular one [3 and 4] and to what extend brand awareness influences the dependent variable. Across all estimations, advertising volume has a positive significant impact on mentions of magazine’s coverage even after controlling for brand awareness and brand popularity [3 and 4], Since AdV is a continuous variable, γ can be interpreted. If the advertising volume increases by one unit, 𝑚𝑚𝑡𝑡𝑡𝑡𝑡𝑡 will differ by γ units on average. Because also brands 19 without any advertising volume were mentioned in the editorial part of the investigated magazines and vice versa, the 𝑚𝑚𝑡𝑡𝑖𝑖𝑖𝑖 - intercept, can be interpreted as the value if a company did not purchased any advertisement. Table 8: Panel poisson regression model (1+2) and panel negative binomial regression model (3+4) per year and company (3) Model 0.000000785** (4) Model 0.000000552** (10.54) (2) Model 0.000000519** * (4.74) (9.36) (4.05) Valuable Brand 1.392*** (5.29) 0.457 (1.16) 1.361*** (5.29) 0.665 (1.64) Brand popularity 1.742*** (8.14) -0.424 (-1.17) 1.686*** (8.31) -0.389 (-1.00) VARIABLES AdVc,t (1) Model 0.000000769*** Brand awareness % _cons Lnalpha _cons ln_r _cons 0.780*** (21.99) -0.459*** (-6.86) 1416 * 0.0251** 0.0227** (3.01) (2.63) 1.560*** (5.70) -0.707** (-3.02) ln_s _cons N * 145 2.173*** (14.10) 2.099*** (4.58) 2.370*** (19.40) 1.821*** (5.50) 0.870*** (9.68) 1416 1.182*** (3.89) 145 t statistics in parentheses * p < 0.05, ** p < 0.01, *** p < 0.001 Table 8 includes the results of the panel poisson regression model (1+2) and panel negative binomial regression model (3+4) per year and company. It does not change the results substantially with regard to advertising volume. Advertising expenditures are significantly correlated to the amount of mentions. Controlling for the 10 most valuable brands for cosmetic products worldwide, the estimations results also turn out to be significant. Current brand awareness has a strong positive impact on the 20 mentions of the representing company as well (see tables 7, 8, 9). Brand value and brand popularity turn out to be insignificant when controlling for brand awareness. To focus on the group-level of companies (concerns), I also ran all of the three model specification for the years 2009-2012. As you can see in table 9, the only considerable difference is that the variable AdV becomes insignificant when panel negative binomial regression model is used (Model 3). Table 9: Linear panel model regression (1), panel poisson regression model (2) and panel negative binomial regression model (3) per year and concern (1) Model 3.16e-05*** (4.04e-06) (2) Model 9.57e-08** (4.24e-08) Valuable Brand 36.86* (20.90) 0.874 (0.725) 0.410 (0.717) Brand popularity -13.58 (20.21) 0.00767 (0.618) -0.496 (0.789) VARIABLES AdVc,t (3) Model 8.43e-08 (5.15e-08) Brand awareness % 0.724* (0.427) 0.0447*** (0.0168) 0.0532*** (0.0178) _cons 1.048* (0.592) -0.00303 (0.288) 1.367** (0.696) lnalpha _cons ln_r _cons ln_s _cons N Number of konz -8.167 (13.14) 69 21 69 21 1.098*** (0.417) 0.349 (0.343) 69 21 Standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 Besides, main problems in the estimation should be unconditionally taken into account. It deals with the problematic of endogeneity caused by missing variables and the problem of reverse causality which leads to an overestimated impact. The proxy for advertising volume can easily create endogeneity. 21 Against this background of reverse causality the question arise, whether the mentions of companies have an impact of the amount of advertising volume in probably future editions. On the other side, mentions in magazines coverage probably increase if the monetary volume of advertising is being purchased by company one or two years before. Table 10 (appendix) shows, that there are hardly any changes in level of significance when including time-variables as a dummy. Awarded prizes As mentioned above, there are also annual prize giving ceremonies for special beauty cosmetic products in both magazines. While the "Prix de Beautés" is annually awarded by an expert jury (dermatologists, industry representatives) in the Cosmopolitan, the annual awarded cosmetic-prize in the Glamour magazines (“Glammy”) is selected by the public readers from a pool of nominated products that were chosen by the editors. To estimate the impact of advertising expenditure on nominess and awarded cosmetic products, it is necessary to control for quality of the nominated and awarded products as well. First of all, I estimated a logit regression model in order to find a relationship between nomination for the “Glammy" and the advertising volume of the year. As you can see in estimation results in table 11 (appendix), the model is statistically significant, what means, that advertising volume also impacts the nomination and winning of a (cosmetic) product in this sample. I also estimated a logit regression model in order to find a relationship between winning the “Prix de Beautés" as a price of special products and the advertising volume. As you can see in table 12 (appendix), the model is statistically significant, what means, that advertising volume also impacts the winning of an award. For estimating the influence of advertising volume, I used the advertising volume of the previous year. The reason behind is that the award presentation of the beauty products is always the beginning of the year. 22 6. Conclusion Existing empirical studies on Media Bias are subjects of controversial discussions in the literature. In this paper, I focused on the so-called commercial Media Bias and investigated how coverage of companies is correlated with the amount of advertising expenditure in woman’s magazines in Germany. To get in line with the debated literature about whether commercial biases, caused by dependent markets, occur, this economic paper presents an empirical analysis of a commercial Media Bias in woman’s magazines. In summary, it can be stated that there is a strong positive correlation between the advertising volume and the number of mentions in the editorial part in this sample. There is also a positive correlation between companies’ nominees, winning an award and their advertising expenditure. Note, however, that this estimation is not perfect, because it is very difficult to measure the quality of the mentioned products in order to control for quality or for objective criterias. Because high-quality products are likely and naturally to be mentioned in magazines’ coverage and high-quality companies are usually more profitable and therefore purchase a higher amount of advertising in the magazines, the regressions would not measure a media bias at all. In order to avoid these problems of endogeneity adequate instruments which are relevant and exogenous for using an instrumental variables technique have to be find (see Wooldridge, 2010). Unfortunately, the problem of endogeneity cannot be easily solved. 23 References Anderson, S. P., Gabszewicz J. J. (2006), The Media and Advertising: a tale of twosided markets". In V. Ginsburgh and D. Throsby (Eds), Handbook of Cultural Economics, Elsevier Science. AWA (2014), Allensbacher Marktanalyse Werbeträgeranalyse, http://www.ifdallensbach.de/fileadmin/AWA/AWA2014/Codebuchausschnitte/AWA2014_Basistabelle_Printmedien.pdf. Bauer Media Group (2014), Cosmopolitan, http://www.baueradvertising.de/cosmopolitan/. Bauer Media Group (2013), Cosmopolitan Preisliste 2013, Nr. 32, gültig ab 01. Januar 2013, www.baueradvertising.de/fileadmin/download/preislisten/2013/ anzeigentarif_cosmopolitan.pdf. Blasco, A., Sobbrio, F. (2011), Competition and Commercial Media Bias, http://www2.dse.unibo.it/wp/WP767.pdf (zuletzt aufgerufen am 12.09.2014). Bagwell, K. (2007), The Economic Analysis of Advertising, in: M. Armstrong and R. Porter (Eds.): Handbook of Industrial Organization, 3, 1701-1844. Brigitte (2010), Kommunikationsanalyse 2010. Frauen in Deutschland: Einstellungen, Marken, Medien, http://ems.guj.de/fileadmin/redaktion/Print/Deutsch/Portfolio/Titel/brigitte/ KA_2010_Berichtsband.pdf. Brigitte (2012), Kommunikationsanalyse 2012. Frauen in Deutschland: Einstellungen, Marken, Medien, http://ems.guj.de/fileadmin/redaktion/Print/Deutsch/Portfolio/Titel/brigitte/KA_2012_Druck_V10.pdf. Cameron, A. C. and Trivedi, P. K. (1998). Regression Analysis of Count Data. New York: Cambridge Press. Chan, J., Suen, W., (2008), A Spatial Theory of News Consumption and Electoral Competition", The Review of Economic Studies, 75(3), 699-728. Condé Nast Verlag (2010), Glamour Ausgabe April 2010, München. Condé Nast Verlag (2012), Glamour Ausgabe Februar 2012, München. Condé Nast Verlag (2014), Glamour. Preisliste 2014, http://www.condenast.de/files/cnv_mk14_glamour_d_3.pdf. Condé Nast Verlag (2014a), Glamour launches, 24 http://www.condenast.com/about-us/heritage/1939/glamour-launches. Dewenter, R. (2006), Das Konzept der zweiseitigen Märkte am Beispiel von Zeitungsmonopolen, Diskussionspapier No. 53. Dewenter, R. (2011), Der Mediensektor zwischen Wettbewerb und Regulierung: aktueller und zukünftiger (De-)Regulierungsbedarf, https://www.wiwi.uni-muenster.de/06/nd/fileadmin/vfs/2011/Dewenter.pdf. Dewenter, R., Heimeshoff, U. (2014), Media bias and advertising: Evidence from a German car magazine. DICE Discussion Paper No. 132. http://hdl.handle.net/10419/93077 (zuletzt aufgerufen am 22.07.2014). Diez, G. (2009), Krise der Printmedien: Wieso die Zeitung in Existenznöten ist, http://www.spiegel.de/kultur/gesellschaft/krise-derprintmedienwieso-die-zeitung-in-existenznoeten-ist-a-623345.html. Ellman, M., Germano, F. (2009), What do the papers sell? A model of advertising and media bias. Economic Journal, 119, 680–704. Gal-Or, E., Geylani, T., and Yildirim, T., P. (2010), The Impact of Advertising on Media Bias", mimeo, University of Pittsburgh. Gambaro, M., Publisi, R. (2009), What Do Ads Buy? Daily Coverage of Listed Companies on the Italian Press, working papers series. Gentzkow M., and Shapiro J. (2006), Media Bias and Reputation, Journal of Political Economy, 114(2), 280-316. Gentzkow, M., Glaeser, E., L., and Goldin, C. (2006), The Rise of the Fourth Estate: How Newspapers Became Informative and Why it Mattered. In Glaeser, E., and Goldin C., (Eds), Corruption and Reform: Lessons from America's Economic History. NBER/University of Chicago Press. Gentzkow, M., Shapiro, J. M. (2010), What Drives Media Slant? Evidence From U.S. Daily Newspapers, Econometrica, 78 (1), 35-71. Guo, W.-C., Lai, F-C (2014), Media Bias When Advertisers Have Bargaining Power, Journal of Media Economics, 27 (3), 120-136. Hamilton, J. T. (2004), All the News that fit to Sell, Princeton: NJ. IVW (2014a), Publikumszeitschriften mit nationaler Verbreitung. MONATLICHE FRAUENZEITSCHRIFTEN. Quartalsauflagen, http://daten.ivw.eu/index.php?menuid=1141&u=&p=&b= 25 le&sv=205&sb=205&t=Publikumszeitschriften+mit+nationaler+Verbreitung&tsub=MONATLICHE+FRAUENZEITSCHRIFTEN. IVW (2014), Publikumszeitschriften mit nationaler Verbreitung. MONATLICHE FRAUENZEITSCHRIFTEN. Quartalsauflagen, http://daten.ivw.eu/index.php?menuid=1141&u=&p= &b=alle&sv=205&sb=205&t=Publikumszeitschriften+mit+nationaler+Verbreitung&tsub=MONATLICHE+FRAUENZEITSCHRIFTEN. Jarren, O. (2010), Die Presse in der Wohlfahrtsfalle. Zur institutionellen Krise der Zeitungsbranche, In: Bartelt-Kircher, Gabriele/Bohrmann, Hans/Haas, Hannes et al.(Hrsg.), Krise der Printmedien: Eine Krise des Journalismus?, Berlin/New York: de Gruyter, 13-31. Kaiser, U., Song, M. (2009), Do media consumers really dislike advertising? An empirical assessment of the role of advertising in print media markets, International Journal of Industrial Organization, 27(2), 292-301. Mullainathan, Sendhil and Shleifer, Andrei (2005), The Market for News, In: American Economic Review, 95(4), 1031-1053. News Aktuell (2008), Werbung steht bei immer mehr Frauen hoch im Kurs, http://www.presseportal.de/pm/6788/1222781/werbung-steht-bei-immermehrfrauen-hoch-im-kurs. Poitras, M., and Sutter, D., (2009), Advertiser pressure and control of the news: The decline of muckraking revisited", Journal of Economic Behavior & Organization, 72, 944-958. Reuter, Jonathan (2009), Does Advertising Bias Product Reviews? An Analysis of Wine Ratings, In: Journal of Wine Economics, 4(2), 125-151. Reuter, J., Zitzewitz, E. (2006), Do Ads Influence Editors? Advertising and Bias in the Financial Media?, In: The Quarterly Journal of Economics, 121 (1), 197-227. Rinallo, D., and Basuroy, S. (2009), Does Advertising Spending Inuence Media Coverage of the Advertiser? Journal of Marketing, 73(6), 33-46. Rochet, J. C., Tirole, J. (2003), Platform Competition in Two-sided Markets, in: Journal ofthe European Economic Association, 1, 990-1029. Sobbrio, F. (2011), A Citizen-Editors Model of News Media", mimeo, IMT Lucca. 26 VDZ/McKinsey (2012), Veränderte Medienlandschaft – Veränderte Media Spendings, http://www.pzonline. de/fileadmin/images/ 121031_VDZ_McKinsey_Studie_Publishers_Summit_final.pdf. VDZ (2014), Anzahl der in Deutschland publizierten Publikumszeitschriften in den Jahren von 1997 bis 2014 [zitiert nach: Statista (2015), Anzahl der in Deutschland publizierten Publikumszeitschriften in den Jahren von 1997 bis 2014. Walter, J. (2014), Media Bias durch Werbung – Das Beispiel Frauenzeitschriften, Masterarbeit TU Ilmenau. Wilhelm-Fischer, H. (2004), Was die neuen Frauen wollen. Einequalitative Studie zum Mediennutzungsverhalten von Leserinnen der Zeitschrift Glamour, Münster: LIT Verlag. Wooldridge, J. (2010), Econometric Analysis of Cross Section and Panel Data, MIT Press, 2. Ed. 27 Appendix Table 10: Linear panel model regression by means of robust estimators VARIABLES AdVc,t (1) Model 1.87e05*** (2.58e06) Valuable Brand 7.751** (3.791) Brand popularity 13.09*** (3.418) 2010. Year 0.456** (0.215) 2011. Year 0.659*** (0.236) 2012. Year 0.589** (0.261) Constant 1.387*** (0.192) 1,416 749 Observations Number of companies Robust standard errors in parentheses *** p<0.01, ** p<0.05, * p<0.1 28 Table 11: Logit regression output for nomination “Glammy” as a dependent variable VARIABLES AdVc,t (1) r1 (2) r2 (3) r3 (4) r4 Model Model Model Model 1.24e-05*** 1.08e-05*** 9.57e-06*** 1.08e-05*** (1.13e-06) (1.16e-06) (2.72e-06) (1.16e-06) 2.532*** -0.352 2.520*** (0.521) (1.059) (0.521) 1.132** (0.530) -1.232 (1.116) 1.122** (0.530) Brand popularity Valuable Brand 2010.Year 0.0127 (0.303) 2011.Year -0.228 (0.308) 2012.Year -0.0161 (0.294) Brand awareness % 0.0388* (0.0213) Constant -2.512*** -2.586*** -2.406*** -2.531*** (0.124) (0.128) (0.733) (0.227) Observations 1,052 1,052 130 1,052 ll df_m -332.1 1 -317.1 3 -55.17 4 -316.7 6 chi2 276.2 306.3 68.38 307.1 Robust standard errors in parenthese *** p<0.01, ** p<0.05, * p<0.1 29 Table 12: Logit regression output for winning the “"Prix de Beautés"” as a dependent Variable VARIABLES AdVc,t (1) r1 Model (2) r2 Model 2.19e-06*** (5.94e-07) 1.77e-06** (7.39e-07) 0.397 (0.561) 0.448 (0.520) -2.009*** (0.239) 243 -111.0 3 16.79 Valuable Brand Brand popularity Constant Observations Ll df_m chi2 -2.004*** (0.237) 243 -111.4 1 15.84 Robust standard errors in parenthese *** p<0.01, ** p<0.05, * p<0.1 30 Diskussionspapiere aus dem Institut für Volkswirtschaftslehre der Technischen Universität Ilmenau Nr. 49 Jaenichen, Sebastian; Steinrücken, Torsten: Opel, Thüringen und das Kaspische Meer, Januar 2006. Nr. 50 Kallfaß, Hermann H.: Räumlicher Wettbewerb zwischen Allgemeinen Krankenhäusern, Februar 2006. Nr. 51 Sickmann, Jörn: Airport Slot Allocation, März 2006. Nr. 52 Kallfaß, Hermann H.; Kuchinke, Björn A.: Die räumliche Marktabgrenzung bei Zusammenschlüssen von Krankenhäusern in den USA und in Deutschland: Eine wettbewerbsökonomische Analyse, April 2006. Nr. 53 Bamberger, Eva; Bielig, Andreas: Mehr Beschäftigung mittels weniger Kündigungsschutz? Ökonomische Analyse der Vereinbarungen des Koalitionsvertrages vom 11. 11. 2005, Juni 2006. Nr. 54 Jaenichen, Sebastian; Steinrücken, Torsten: Zur Ökonomik von Steuergeschenken - Der Zeitverlauf als Erklärungsansatz für die effektive steuerliche Belastung, Dezember 2006. Nr. 55 Jaenichen, Sebastian; Steinrücken, Torsten: Wirkt eine Preisregulierung nur auf den Preis? Anmerkungen zu den Wirkungen einer Preisregulierung auf das Werbevolumen, Mai 2007. Nr. 56 Kuchinke, B. A.; Sauerland, D.; Wübker, A.: Determinanten der Wartezeit auf einen Behandlungstermin in deutschen Krankenhäusern - Ergebnisse einer Auswertung neuer Daten, Februar 2008. Nr. 57 Wegehenkel, Lothar; Walterscheid, Heike: Rechtsstruktur und Evolution von Wirtschaftssystemen - Pfadabhängigkeit in Richtung Zentralisierung?, Februar 2008. Nr. 58 Steinrücken, Torsten; Jaenichen, Sebastian: Regulierung und Wohlfahrt in einem Modell mit zwei Aktionsparametern, März 2008. Nr. 59 Lehnert, Ninja M.: Externe Kosten des Luftverkehrs - Ein Überblick über den aktuellen Stand der Diskussion, April 2008. Nr. 60 Walterscheid, Heike: Reformbedarf etablierter Demokratien im Kontext dezentralisierter Gesellschaftssysteme - Grundlegende Hindernisse bei Steuersystemreformen“, April 2010. 31 Nr. 61 Walterscheid, Heike; Wegehenkel, Lothar: Kostenstruktur, Zahlungsbereitschaft und das Angebot von Mediengütern auf Medienmärkten, Juni 2008. Nr. 62 Walterscheid, Heike; Wegehenkel, Lothar: Wohlstand der Nationen und handlungsrechtliche Struktur eines Gesellschaftssystems, September 2008. Nr. 63 Dewenter, Ralf; Haucap, Justus; Wenzel, Tobias: Indirect Network Effects with Two Salop Circles: The Example oft the Music Industry, Juni 2009. Nr. 64 Dewenter, Ralf; Jaschinski, Thomas; Wiese, Nadine: Wettbewerbliche Auswirkungen eines nichtneutralen Internets, Juli 2009. Nr. 65 Dewenter, Ralf; Haucap, Justus; Kuchinke, Björn A.: Das Glück und Unglück von Studierenden aus Ost- und Westdeutschland: Ergebnisse einer Befragung in Ilmenau, Bochum und Hamburg, Oktober 2009. Nr. 66 Kuchinke, Björn A.; Zerth, Jürgen; Wiese, Nadine: Spatial Competition between Health Care Providers: Effects of Standardization, Oktober 2009. Nr. 67 Itzenplitz, Anja; Seifferth-Schmidt, Nicole: Warum Klimakonferenzen scheitern, aber dennoch zum Wohl des Weltklimas kooperiert wird, Juli 2010. Nr. 68 Kallfaß, Hermann H.: Die Aufmerksamkeit für, die Nutzung der und die Werbung in Medien in Deutschland, November 2010. Nr. 69 Budzinski, Oliver: Empirische Ex-Post Evaluation von wettbewerbspolitischen Entscheidungen: Methodische Anmerkungen, Januar 2012. Nr. 70 Budzinski, Oliver: The Institutional Framework for Doing Sports Business: Principles of EU Competition Policy in Sports Markets, January 2012. Nr. 71 Budzinski, Oliver; Monostori, Katalin: Intellectual Property Rights and the WTO, April 2012. Nr. 72 Budzinski, Oliver: International Antitrust Institutions, Juli 2012. Nr. 73 Lindstädt, Nadine; Budzinski, Oliver: Newspaper vs. Online Advertising - Is There a Niche for Newspapers in Modern Advertising Markets? Nr. 74 Budzinski, Oliver; Lindstädt, Nadine: Newspaper and Internet Display Advertising - Co-Existence or Substitution?, Juli 2012b. Nr. 75 Budzinski, Oliver: Impact Evaluation of Merger Control Decisions, August 2012. 32 Nr. 76 Budzinski, Oliver; Kuchinke, Björn A.: Deal or No Deal? Consensual Arrangements as an Instrument of European Competition Policy, August 2012. Nr. 77 Pawlowski, Tim, Budzinski, Oliver: The (Monetary) Value of Competitive Balance for Sport Consumers, Oktober 2012. Nr. 78 Budzinski, Oliver: Würde eine unabhängige europäische Wettbewerbsbehörde eine bessere Wettbewerbspolitik machen?, November 2012. Nr. 79 Budzinski, Oliver; Monostori, Katalin; Pannicke, Julia: Der Schutz geistiger Eigentumsrechte in der Welthandelsorganisation - Urheberrechte im TRIPS Abkommen und die digitale Herausforderung, November 2012. Nr. 80 Beigi, Maryam H. A.; Budzinski, Oliver: On the Use of Event Studies to Evaluate Economic Policy Decisions: A Note of Caution, Dezember 2012. Nr. 81 Budzinski, Oliver; Beigi, Maryam H. A.: Competition Policy Agendas for Industrializing Countries, Mai 2013. Nr. 82 Budzinski, Oliver; Müller, Anika: Finanzregulierung und internationale Wettbewerbsfähigkeit: der Fall Deutsche Bundesliga, Mai 2013. Nr. 83 Doose, Anna Maria: Methods for Calculating Cartel Damages: A Survey, Dezember 2013. Nr. 84 Pawlowski, Tim; Budzinski, Oliver: Competitive Balance and Attention Level Effects: Theore-tical Considerations and Preliminary Evidence, März 2014. Nr. 85 Budzinski, Oliver: The Competition Economics of Financial Fair Play, März 2014. Nr. 86 Budzinski, Oliver; Szymanski, Stefan: Are Restrictions of Competition by Sports Associations Horizontal or Vertical in Nature?, März, 2014. Nr. 87 Budzinski, Oliver: Lead Jurisdiction Concepts Towards Rationalizing Multiple Competition Policy Enforcement Procedures, Juni 2014. Nr. 88 Budzinski, Oliver: Bemerkungen zur ökonomischen Analyse von Sicherheit, August 2014. Nr. 89 Budzinski, Oliver; Pawlowski, Tim: The Behavioural Economics of Competitive Balance: Implications for League Policy and Championship Management, September 2014. 33 Nr. 90 Grebel, Thomas; Stuetzer, Michael: Assessment of the Environmental Performance of European Countries over Time: Addressing the Role of Carbon, September 2014. Nr. 91 Emam, Sherief; Grebel, Thomas: Rising Energy Prices and Advances in Renewable Energy Technologies, July 2014. Nr. 92 Budzinski, Oliver; Pannicke, Julia: Culturally-Biased Voting in the Eurovision Song Contest: Do National Contests Differ?, December 2014. Nr. 93 Budzinski, Oliver; Eckert, Sandra: Wettbewerb und Regulierung, März 2015. Nr. 94 Budzinski, Oliver; Feddersen, Arne: Grundlagen der Sportnachfrage: Theorie und Empirie der Einflussfaktoren auf die Zuschauernachfrage, Mai 2015. Nr. 95 Pannicke, Julia: Abstimmungsverhalten im Bundesvision Song Contest: Regionale Nähe versus Qualität der Musik, Oktober 2015. Nr. 96 Budzinski, Oliver; Kretschmer, Jürgen-Peter: Unprofitable Horizontal Mergers, External Effects, and Welfare, October 2015. Nr. 97 Budzinski, Oliver; Köhler, Karoline Henrike: Is Amazon The Next Google?, October 2015. Nr. 98 Kaimann, Daniel; Pannicke, Julia: Movie success in a genre specific contest: Evidence from the US film industry, December 2015. 34